Intrafractional Prostate Motion and Dose Variation during Radiation Therapy for Prostate Cancer Using 4D-TPUS

Abstract

More recently, noninvasive 4D transperineal ultrasound (4D-TPUS) has been introduced in tracking intrafractional prostate motion in radiotherapy. Compared to other tracking method, the ultrasound has its own advantage in precise identification of soft tissue without invasive procedure or extra radiation dose. In this study, we investigated the prostate motion model and dose variation during radiotherapy using a transperineal autoscan ultrasound system (Clarity®, Elekta Inc., Sweden). From July 2014 through December 2016, 57 prostate cancer patients were enrolled with a median age of 74 years (range 61-85). All patients were asked to have a full bladder and empty rectum before each treatment. A simultaneous CT and TPUS simulation was conducted to provide a fused CT/US reference. Before radiation delivery, prostates were localized using CBCT to determine setup offsets relative to the patients’ skin tattoos. During the treatment, real-time ultrasound images were acquired and data was collected for direct monitoring of 3D motion of the prostate. Then dose distributions were recalculated with the isocenter shifts relative to the simulation CT images according to the real-time data using the leaf sequences/MUs based on the original treatment plan. The doses were compared with the original doses planned on the simulation CT using the clinical acceptance criteria. A total of 1207 fractions were evaluated. The mean (±SD) of the infraction displacements were [mm]: I(+)/S: (0.03 ± 0.92); L(+)/R: (0.12 ± 1.03); and A(+)/P: (-0.01 ± 1.37), respectively. There were 41/1207 (3.4%), 29/1207 (2.4%), and 66/1207 (5.5%) fractions with deviation exceeded 3 mm in the IS, LR, and AP directions, respectively. The A/P direction has the largest extent of prostate displacement. The percentage of time with displacements 0-1mm, 1-2mm, 2-3mm and larger than 3 mm was 88.1%, 9.8%, 1.7% and 0.4% in the IS direction, 89.3%, 8.9%, 1.5% and 0.3 % in the LR direction, 84.5%, 11.1%, 3.7%, and 0.7% in the AP direction. We classified our patients into three groups according to the motion model: stable (n=50), irregular (n=3) and intention (n=4). The intention group was defined as persistent deviation to the same direction (the maximal displacement should exceed 3 mm) that repeated in at least 50% fractions. All these patients had obviously anxiety and urinary frequency and/or urgency. Recalculated dose distributions showed that all the 4 patients in the intention group did not meet our criterion of D95% of PTV prostate> 8000 cGy or V70 of rectum< 20%, which may leading to insufficient treatment and increased toxicity. The present study demonstrated that for over 90% of fractions, a CTV-PTV margin of 3 mm would be good coverage with the planned prescribed dose. However, it is important to identify the patients belonging to the intention group. Pre-treatment with anti-anxiety drugs and α receptor blocker may be useful in relief the prostate motion.